Failure prediction of European high-tech companies

Abstract

The aim of this thesis is to develop a model for predicting the failure of high-tech and mediumhigh tech companies from different European countries. This study uses firm-level data from the Bureau van Dijk’s Amadeus database and includes the financial information of 32,929 firms. The data were collected from the financial statements of the companies for the period 2012–2017 and logistic regression was used as the analysis method. Findings indicate that the accuracies of individual variables across countries are not very high and there are large differences in the accuracies of individual ratios when comparing non-failed and failed firms. Aggregate accuracies for all ratios within country and across countries show that the most accurate predictions are obtained for non-failed firms using the ratios for the preceding two years combined. The practical value of this work lies in the knowledge of the relevant variables, which allows companies to focus in a timely manner on aspects that have determined failure in the past. Subsequent works should attempt to use a larger sample of European countries and include other variables in addition to financial ratios.